Any updates since 1799552?

[sa...@dunquino.com]

Hi,

I noticed that my channels updates cf file hasn't changed for a month so I ran sa-update with --debug:

Oct 23 09:58:35.372 [4969] dbg: channel: attempting channel updates.spamassassin.org
Oct 23 09:58:35.372 [4969] dbg: channel: using existing directory /var/opt/spamassassin/3.004001/updates_spamassassin_org
Oct 23 09:58:35.372 [4969] dbg: channel: channel cf file /var/opt/spamassassin/3.004001/updates_spamassassin_org.cf
Oct 23 09:58:35.372 [4969] dbg: channel: channel pre file /var/opt/spamassassin/3.004001/updates_spamassassin_org.pre
Oct 23 09:58:35.373 [4969] dbg: channel: metadata version = 1799552, from file /var/opt/spamassassin/3.004001/updates_spamassassin_org.cf
DNS TXT query: 1.4.3.updates.spamassassin.org -> 1799552
Oct 23 09:58:35.378 [4969] dbg: dns: 1.4.3.updates.spamassassin.org => 1799552, parsed as 1799552
Oct 23 09:58:35.378 [4969] dbg: channel: current version is 1799552, new version is 1799552, skipping channel
Oct 23 09:58:35.378 [4969] dbg: diag: updates complete, exiting with code 1
Update finished, no fresh updates were available

I get this with dig:

$ dig -t txt 1.4.3.updates.spamassassin.org +short
3.3.3.updates.spamassassin.org.
"1799552"

So it looks like 1799552 really is the latest version. What concerns me though is that this revision is mentioned in the mailing list back in July.

Is 1799552 really the latest set of rules for v3.4.1?

Thanks,
orval

Re: Any updates since 1799552?

[David Jones <dj...@ena.com>]

On 10/23/2017 05:03 AM, sa_list@dunquino.com wrote:
> Hi,
> 
> I noticed that my channels updates cf file hasn't changed for a month so I ran sa-update with --debug:
> 
> Oct 23 09:58:35.372 [4969] dbg: channel: attempting channel updates.spamassassin.org
> Oct 23 09:58:35.372 [4969] dbg: channel: using existing directory /var/opt/spamassassin/3.004001/updates_spamassassin_org
> Oct 23 09:58:35.372 [4969] dbg: channel: channel cf file /var/opt/spamassassin/3.004001/updates_spamassassin_org.cf
> Oct 23 09:58:35.372 [4969] dbg: channel: channel pre file /var/opt/spamassassin/3.004001/updates_spamassassin_org.pre
> Oct 23 09:58:35.373 [4969] dbg: channel: metadata version = 1799552, from file /var/opt/spamassassin/3.004001/updates_spamassassin_org.cf
> DNS TXT query: 1.4.3.updates.spamassassin.org -> 1799552
> Oct 23 09:58:35.378 [4969] dbg: dns: 1.4.3.updates.spamassassin.org => 1799552, parsed as 1799552
> Oct 23 09:58:35.378 [4969] dbg: channel: current version is 1799552, new version is 1799552, skipping channel
> Oct 23 09:58:35.378 [4969] dbg: diag: updates complete, exiting with code 1
> Update finished, no fresh updates were available
> 
> I get this with dig:
> 
> $ dig -t txt 1.4.3.updates.spamassassin.org +short
> 3.3.3.updates.spamassassin.org.
> "1799552"
> 
> So it looks like 1799552 really is the latest version. What concerns me though is that this revision is mentioned in the mailing list back in July.
> 
> Is 1799552 really the latest set of rules for v3.4.1?
> 
> Thanks,
> orval
> 

Yes.  That is the latest.  There is a very odd bug in the nightly 
masscheck processing that validates the latest ruleset.  The 
72_scores.cf file that it generates was about 1/3rd the size it should 
be.  This caused a number of scores to take the default score of 1.0 and 
much lower overall scoring.  Many people noticed their SA was scoring 
very low and spam was getting through for a couple of weeks in late June.

There haven't been any major updates or additions to rules, scores, or 
plugins the past 6 months so we really haven't missed much.  There could 
be some new RBLs added to the default rules with low scores for a slow 
introduction plus a few very minor updates coming soon so the updates 
should start rolling again soon after a few days of testing.

SA installations running mostly default rules and scores are more 
depending on the nightly score updates in 72_scores.cf than heavily 
customized installations.  My installation has a lot of local additions 
of RBLs, custom meta rules, some increased scores, etc. with a heavily 
tuned MTA in front of SA.  I noticed a slight drop in scores in late 
June but I didn't have major problems with spam getting through.  I was 
watching closely and didn't even notice the problem for two weeks until 
I saw reports on the MailScanner mailing list.

-- 
David Jones